One Image HDR ???

[scorpio_e]

I really thought I knew this but I am beginning to question what I think I know.

If you have a single image and tone map it in PhotoMatrix, can you consider it HDR?

[kat]

I always thought with one image that one still had to get at least another image with different exposure to do it. Changing the second photos exposure using PS or whatever it is that you use. I may be wrong..haven't done it that way. Hmmm..good question

[scorpio_e]

I understand what you mean. Use one raw file.. vary the exposure and then bring the three images in PS or Photomatrix.

My quest is... ONE single image.. bring the single image into PS or Photomatrix..Next step..tone map..

Would it be HDR or not?



I was REALLY convinced that is it not but now..HMmmmm..I guess I really do not understand what tone mapping does on a single image...

[scorpio_e]

Nice blog Kat:)

[Richard]

the RAW file contains more a wider dynamic range than can be saved in a Jpg format. The tone mapping in simple terms will compress this dynamic range into a displayable image. A one shot HDR could be considered an expanded range image. If you want a true high dynamic range you need to capture multiple bracketed images. Different camera capture different ranges within a single image, last time I check Pentax wear slightly ahead of Canon and Nikon.

[Mad Aussie]

I think it's called a Faux HDR in the case of a single image being reproduced and exposure changed. I've done it before as a test in Photomatix but I do it more often in Photoshop and use manual techniques, no Photomatix etc.

[scorpio_e]

I think it's called a Faux HDR in the case of a single image being reproduced and exposure changed. I've done it before as a test in Photomatix but I do it more often in Photoshop and use manual techniques, no Photomatix etc.

So what if you do not change the exposure and just tone map? Would you consider it HDR or Faux HDR?

[Mad Aussie]

I think I'd have to try it but I don't think Photomatix lets you do that without adjusting the exposure values.

[Rinkdaddy]

I use HDR Efex Pro by Nik Software as a plugin for Aperture and it allows me to bring in a single raw image and use all the presets and adjustments to fine tune it. Normally I use either 3 or 5 images but have used only one if something isn't lining up right or there happens to be dust or water marks on my lens and then it gets magnified when I use HDR with more than one image. Thank goodness for the retouch tool. I have to get used to checking the lens for dust when I shoot for HDR especially for the sky where it really shows up.

[taffy]

I think they call it pseudo-HDR when you only use a single file. This is cut and paste from the Photomatix manual.

"Photomatix Pro allows you to create a 32-bit HDR image from a single RAW file.To do this, open one RAW file using File >Open, and Photomatix will convert it into a pseudo-HDR image. It is important to note, though, that an image created with a single RAW file cannot really be considered High Dynamic Range. The important characteristic of this pseudo-HDR image is that it is unprocessed. Its dynamic range is not much larger than the range of an already converted file."

I have never got the same effect when I have used a single file compared to using bracketed files.

Taffy:)

[Barefoot]

EDR: Extended Dynamic Range.

[Marko]

i don't think it matters what it's called actually....but it is totally doable...
Taffy is correct, bracketed exposures are better than a pseudo HDR head to head - but you can still achieve A LOT from just one file.

[Andrew]

With one shot you'd need a Lytro or Raytrix plenoptic camera. If you haven't seen the first video from Adobe watch it. Amazing.

Adobe Plenoptic Demo from NVIDIA GTC 2010 - YouTube (http://www.youtube.com/watch?v=-EI75wPL0nU)

Lytro - The Start of a Picture Revolution - YouTube (http://www.youtube.com/watch?v=7babcK2GH3I)

[ericmark]

I think one does have to define HDR the Fuji's SuperCCD S3 Pro camera has a primary and secondary cells allowing a far greater range to be captured than with cameras without the special CCD and clearly can produce a HDR picture with one image. But when I try and I guess most cameras are the same to lighten the dark areas of a RAW image I get noise or grain which ever name you want to give it which limits how much I can lighten dark areas, and with whited out areas sorry nothing can be done.

So using the RAW tools be it a S curve, spot adjustment, or fill light you can compact more into a Jpeg image then it would normally have straight out out of the camera but as to if this is HDR first one must define how many EV stops is HDR. http://www.cybergrain.com/tech/hdr/images1/finepix.jpg With the sensor shown it may need no software to be considered HDR.

What we want is to compact the dynamic range to what can be presented by the display medium. As well as HDR programs one can also use layers and masks often quicker and with better results. Where the range is spread throughout the image HDR software is often the only way but where for example a window shows a view outside but you also want the details of inside so you have a reasonably well defined change using layers and masks works far better as it does not reduce the level with either version of the images. Also often quicker too.

With Jpeg cameras often setting the camera to bracket the shots will allow between 1/3 to 1 EV between each shot where with a camera taking in RAW this is often as big as 2 EV between shots so one can add with 5 images between 5 and 10 EV to the image according to if taken in Jpeg or RAW. Both are called HDR even though one has much more range then the other. It was called high exposure latitude and again no one defined it.

[Gremlich]

Here's a one image HDR

14122

[ericmark]

Due to the effects included it is hard to see the HDR content. The sky has a lot of grain which seems wrong way around normally it's the dark areas. Many programs include warping software of some sort which can produce some really odd effects. This can also be done by including a reversal in part of the curve. The fact that some HDR software includes these warping tools does not make a picture which has been warped into a HDR image.

There is not to my knowledge a point at which an image is considered as HDR but it is all to do with the EV range of the image. Different cameras will capture different amounts and may include some software to compact the range. We can use 8, 16, or 32 bit to capture and store images my camera uses 16 bit and as a result has more information stored than with a cheaper camera which records images as 8 bit. But to use the extra information needs some manual intervention to select which bits to compact. (Tone mapping) I will normally use RAW 5.7 from Photoshop but there was software also provided by Pentax which is not as easy to use but does the same thing.

If we use one of the more basic programs like UFRaw (Used with Gimp to handle RAW files) we will see far more sliders and less auto tools like fill light and recovery and no adjustment brush. Which means much harder to maintain the detail required when converting from 16 to 8 bit.

Using both Photoshop and Picturenaut to combine images we end up with a 32 bit image and just like when we use RAW 5.7 to convert a 16 bit to 8 bit we have to use software to convert 32 bit to 16 or 8 bit. It does not really matter if it's called "Convert to HDR" as with Photoshop or "Tone Mapping" as with Picturenaut we have to select how to reduce the information to squeeze it into 8 bits.

So with a standard 8 bit image we could of course do the same. In Photoshop if we go to image/adjustments/curves and place some troughs in the curve we will get some really weird effects. This is nothing to do with being a HDR image. In photoshop we have hundreds of options like emboss and by making a collection of images with different effects then combining with layers and masks we can completely transform the image. But this is not HDR.

If you were to however take a RAW image and convert it twice into a Jpeg once under exposed and once over exposed then take the two images and combine them using layers and masks so that the dark bits are masked in dark version and light bits masked in light version then yes you have produced a HDR image from a single image.

To me I consider over 10 EV to be HDR but there does not seem to be a standard figure. It's the same with High Definition my monitor is not classed as HD but it has a greater definition than my TV which is classed as HD. At the moment High Fidelity in radio is becoming a question in the UK with the BBC reducing the range used with DAB transmissions so one can get better sound on TV.

There are some fixed points. Low voltage with AC is between 50 - 1000 volts so high voltage is over a 1000 volts. However most people would consider the 230v found in European homes as being high volts. It seems this word "High" can be very misleading.

I am lead to believe some of the large format cameras can capture 12 EV stops on a single image so may be HDR starts at 14 EV stops? I have seen some posts referring to 21 EV stops. But clearly the H in HRD refers to a high EV number not a High Warp Factor. Maybe we should have HWR photos with a High Warp Range?