LaserSoft Imaging

Hi.

I'm sorry if this has been answered before. I couldn't find any place where straight answers to my questions were available.

I have AI Studio (AFL, Reflecta RPS 7200) plus HDR Studio. From what I've understood, my best possible workflow for negative archival would be:

- scan to 64-bit HDRi
- with maximum optical resolution
- write as many EXIF data as I'll want later on
- no cropping (it's a film scanner processing a 24.3 x 36.5 mm area at a time)
- no corrections

- open the scans in HDR Studio someday, do all processing and save the result to printable files


My questions are:

- At what point should I apply negafix? My instinct is that the correct procedure would be to scan as positive, and only apply negafix within HDR Studio. But I wonder if the correct selection of negafix before scanning wouldn't help the scanning process? I'm unaware of the specifics of the scanning; whether Silverfast operates the scanner in different ways or not according to whether it's in positive or negative mode, for instance. In particular, I'm worried that scanning a negative in positive mode is somehow sub-optimal (because of the orange mask, if nothing else). But maybe this is a completely wrong idea. I've also heard about scanning with 'linear' negafix, is that what I should do? And only later apply the right negafix within HDR Studio?

- Do I lose anything by applying iSRD only in HDR Studio? I know that the great part of ICE is that the scanner reads the IR data and so is aware of where the dust and scratches are. What I'm wondering is whether there is any other part of the iSRD process that depends on hardware and has to be done during the scan, and can only be approximated if done by software later on. I'm assuming no - that reading the IR channel first and applying iSRD later on gives equal (or better, if the algorithms improve with time) results to simply applying iSRD during the scan time.

- Speaking of iSRD, I've heard that Photoshop CS5 will have very improved scratch-removal algorithms. Is there any chance similar algorithms (if they exist) can be implemented in Silverfast, or that Photoshop can be made IR-aware, to apply them to files produced by HDR Studio?

- Is there any tool to edit EXIF data in a file? I mean, I'd like to the file not to be changed, only the EXIF data. Imagine if I had forgotten something in the scannign step, would there be a way to add it later on without any extra processing?

- I've heard that EXIF data is not always well-interpreted by programs and some may even 'read' it as noise into the image. Should I worry about that? EXIF is not vital to me - it's just that I'm thinking that I intend to scan each image only one, but may process it multiple times in HDR Studio - and inputting EXIF right into the original scan would propagate it easily into any derived image (I suppose).

- Will I have any problem because of scanning the whole scanner area (24.3 x 36.5 mm) with no cropping? I know that image-processing algorithms only work OK if only the usable parts of the image are used. But since I intend to do all processing in HDR Studio, I can *then* process a usable crop of the image. In those conditions, would I have any problem in doing the scan with the whole area, film borders included? IINM, other programs have an option to 'ignore' a certain margin of the frame when applying image algorithms (say, they'll only use 23x35 mm to do the calculations, and then they apply the treatment to the whole 24x36). Does Silverfast consider having a similar option? Maybe the drawable frame border could include a shaded inset to indicate the margin that wouldn't be considered for calculations.

I'm sorry for all these questions, but I couldn't find definite answers for them.
Thank you very much,
Antonio

Hi Antonio.

Antonio Marques wrote:I'm sorry if this has been answered before. I couldn't find any place where straight answers to my questions were available.

All your questions have been answered (not all in one place but...) in our documentation, website and video tutorials, so I will try to address them briefly.

Basically you gotta have in mind that using 48 bit HDR or 64 bit HDRi, generate raw files which are non-modified files containing the complete information delivered by the scanner

Antonio Marques wrote:- At what point should I apply negafix? My instinct is that the correct procedure would be to scan as positive, and only apply negafix within HDR Studio. But I wonder if the correct selection of negafix before scanning wouldn't help the scanning process? I'm unaware of the specifics of the scanning; whether Silverfast operates the scanner in different ways or not according to whether it's in positive or negative mode, for instance. In particular, I'm worried that scanning a negative in positive mode is somehow sub-optimal (because of the orange mask, if nothing else). But maybe this is a completely wrong idea. I've also heard about scanning with 'linear' negafix, is that what I should do? And only later apply the right negafix within HDR Studio?

It does not matter what profile you select during the scan because it is only used for previsualization purposes.

Antonio Marques wrote:- Do I lose anything by applying iSRD only in HDR Studio? I know that the great part of ICE is that the scanner reads the IR data and so is aware of where the dust and scratches are. What I'm wondering is whether there is any other part of the iSRD process that depends on hardware and has to be done during the scan, and can only be approximated if done by software later on. I'm assuming no - that reading the IR channel first and applying iSRD later on gives equal (or better, if the algorithms improve with time) results to simply applying iSRD during the scan time.

When you create a 64 bit HDRi file, you are saving the information read by the infrared channel into the file which will be used by HDR Studio to make the infrared correction.

Antonio Marques wrote:- Speaking of iSRD, I've heard that Photoshop CS5 will have very improved scratch-removal algorithms. Is there any chance similar algorithms (if they exist) can be implemented in Silverfast, or that Photoshop can be made IR-aware, to apply them to files produced by HDR Studio?

Please do not get me wrong, I think you are going too fast on this one, Photoshop CS5 is not yet out but even if it were, it is very likely those algorithms will not be disclosed by adobe, or.. perhaps those are similar to those already used by iSRD (impossible to tell). Photoshop can not read the infrared information contained in the 64 bit HDRi files and it might also destroy the infrared channel (it already happens with ACDsee) rendering your files unusable even by HDR Studio.

Antonio Marques wrote:Is there any tool to edit EXIF data in a file? I mean, I'd like to the file not to be changed, only the EXIF data. Imagine if I had forgotten something in the scannign step, would there be a way to add it later on without any extra processing?

You can use the IPTC buttons included in Ai Studio and HDR Studio to write the desired metadata.

Antonio Marques wrote:Will I have any problem because of scanning the whole scanner area (24.3 x 36.5 mm) with no cropping? I know that image-processing algorithms only work OK if only the usable parts of the image are used. But since I intend to do all processing in HDR Studio, I can *then* process a usable crop of the image. In those conditions, would I have any problem in doing the scan with the whole area, film borders included? IINM, other programs have an option to 'ignore' a certain margin of the frame when applying image algorithms (say, they'll only use 23x35 mm to do the calculations, and then they apply the treatment to the whole 24x36). Does Silverfast consider having a similar option? Maybe the drawable frame border could include a shaded inset to indicate the margin that wouldn't be considered for calculations.

No problem, you can later frame the image on HDR.

Cheers

Hi,

Yes, I've seen answers to those questions here and there. What I hadn't seen yet were absolute statements as to the One True Result of doing this and that. Now from your answers I hope I can gather the following:

1. Scanning to HDR(i) really results in 'raw' files with no processing apart from the combination of multi-exposure results. So it's indifferent whether one works in positive or negative mode, what negafix profile is selected, the applied gamma, colour corrections, exposure corrections, or the amount of the frame that is scanned - the resulting HDR file will be exactly the same because those are all post-scan processing that doesn't get written to HDR and can be done in HDR Studio later on.

2. iSRD is also purely software applied to the IR channel data, so there is never anything lost by applying it later on, instead of during the scan.

3. When processing the image in HDR Studio, one must be careful to select only those parts of the image that have good image data. A border or a poorly-exposed margin can skew the algorithms' operation. This is where I think a nice feature could be added (if it doesn't exist already): allow to specify a certain margin (in percent or pixels) to be ignored when doing calculations, but to which the corrections will be applied as well and will be present in the processed image.

4. Currently no program other than HDR Studio is able to process HDRi files. I think it would be very nice if the HDRi format (or some variant of it) became a standard that other programs could work with. Are there any plans to make it so? Is it a standard extension to TIFF that everyone can implement if willing to take the effort?

Thanks.
Antonio

Hi Antonio,

Antonio Marques wrote:1. Scanning to HDR(i) really results in 'raw' files with no processing apart from the combination of multi-exposure results. So it's indifferent whether one works in positive or negative mode, what negafix profile is selected, the applied gamma, colour corrections, exposure corrections, or the amount of the frame that is scanned - the resulting HDR file will be exactly the same because those are all post-scan processing that doesn't get written to HDR and can be done in HDR Studio later on.

That is absolutely right

Antonio Marques wrote:2. iSRD is also purely software applied to the IR channel data, so there is never anything lost by applying it later on, instead of during the scan.

iSRD means the infrared channel will be used to make corrections, in case of the HDRi files the infrared data is stored in one channel and used latter to perform the corrections that would otherwise have been done during the scan process.

Antonio Marques wrote:3. When processing the image in HDR Studio, one must be careful to select only those parts of the image that have good image data. A border or a poorly-exposed margin can skew the algorithms' operation. This is where I think a nice feature could be added (if it doesn't exist already): allow to specify a certain margin (in percent or pixels) to be ignored when doing calculations, but to which the corrections will be applied as well and will be present in the processed image.

Check this out: http://www.youtube.com/watch?v=TLyudHy-vlI

Antonio Marques wrote:4. Currently no program other than HDR Studio is able to process HDRi files. I think it would be very nice if the HDRi format (or some variant of it) became a standard that other programs could work with. Are there any plans to make it so? Is it a standard extension to TIFF that everyone can implement if willing to take the effort?

As of today, there are no plans to make it an open file format.

Cheers

LSI_Morales wrote:Check this out: http://www.youtube.com/watch?v=TLyudHy-vlI

Thanks! That was very helpful because I'm trying to produce scans that show the black film border, but I don't want the black film border to influence the exposure of the image.

Just one question: In the video, he was performing those steps during the scanning stage (using SiverFast Ai). But, since I only use SilverFast Ai to create raw 48-bit HDR TIFF negatives, I'd ONLY need to perform those steps he performed in the video while I'm using HDR Studio (after the scanning stage), right?

I'm assuming that I don't need to follow his steps during the scanning stage since I'm just creating 48-bit HDR TIFF negatives during my initial scans. Those initial 48-bit HDR scans aren't influencing the exposure of the negative, right? Or, am I mistaken? Thanks!

Brandon

Hi Brandon,

Brandon D. wrote:n the video, he was performing those steps during the scanning stage (using SiverFast Ai). But, since I only use SilverFast Ai to create raw 48-bit HDR TIFF negatives, I'd ONLY need to perform those steps he performed in the video while I'm using HDR Studio (after the scanning stage), right?

I'm assuming that I don't need to follow his steps during the scanning stage since I'm just creating 48-bit HDR TIFF negatives during my initial scans. Those initial 48-bit HDR scans aren't influencing the exposure of the negative, right? Or, am I mistaken? Thanks!

You are absolutely right, you do not need to perform those steps during the scanning process because you are creating raw files anyway. You can do these during the editing process in SilverFast HDR

Cheers