Just gave Photomatix a try, and thought post the result.
The image created from 3 raw V-Ray renders, each with different exposure.
Raw V-ray render image (1 out of 3) Img No’ 1
Photomatix outcome Img No’ 2
Lot of detail achieved compared to the raw image. I don’t know if it’s the perfect scene to show the advantage of this, but i like it.
I think i’ll use it in my next project.
Well the images in the photomatix gallery are amazing. Doesnt photoshop do this automatically as well? Just cuious because we have photoshop here but I have been using GIMP so much lately I have lost track of all its features.
40th…Great experiment…I’ve actually been meaning to play around with doing some Local tone mapping from HDR renderings from V-Ray. The only thing I’ll say is that you shouldn’t have to render 3 times to get a image that you can use for an HDR->LDR conversion. In photography, yes you do because you’re limited to 8 bit color info usually, and also in that cameras will clip bright pixels if they are too over exposed. With a rendering from V-Ray, you have all of the information that could possibly be there regardless of the exposure you render it at, so there’s no need to combine exposures because one image already has all of that information. One of the other reasons why you combine multiple exposures is to reduce the noise in the image. Since you control the noise in your rendering, there’s no need to combine different exposure’s just to reduce it.
It’s very simple actually.
1. Use physical camera
2. Render a normal exposure image
3. Render Over-exposed image after decreasing Shutter speed
4. Render Under-exposed image increasing Shutter speed
5. Generate HDR image via Photomatix (Here’s the tutorial)
What kind of image are we talking about here? Because i’ve tried to make HDR looking image out of one render image (jpg), by making two additional images via PS (changing the exposure). And the result is worse than the 3 images combination.
Here it is Img No’ 3
There has to be a better way of doing this then rendering 3 separate images. Sure, it would be better for reducing noise like damien said but will it work the same if you adjust the exposure from the Vray Frame buffer or do you actually have to adjust the shutter speed? Ill have to admit, those images look beautiful.
40th…I think your mixing up separate parts of the process. There are two components to what photomatrix is doing. 1) combining multiple exposures of an image into an HDR file, and 2) adjusting that HDR image so that all of data lies within the 8 bit color range.
The first part is a necessity if you’re shooting photos in the the real world for several reasons.
First off, camera sensors will only be able to retain a certain amount of information from the environment. Secondly, you cannot capture into a high dynamic range format. Using Raw files do help, and will get you access to more range than just a jpeg, but ultimately, its not nearly enough. The only solution at the moment is to combine multiple exposures to get the details of all areas of the scene without them either being clipped from being too dark to “excite” the sensor or being so bright that it goes beyond what the sensor can handle.
The second part of the process, referred too as tone mapping, takes that high dynamic range data, which lies beyond what our devices can actually display (or print), and remaps that data back into the range that be displayed and/or printed. The only precursor to this process is that you have an HDR image to start.
In V-Ray, you’re not limited to a range that your image can contain, provided that you A) save as either an HDR or EXR file and B) the you haven’t explicitly clamped the image. Therefore, any exposure should have all of the information that you need to tone map an HDR image back to an LDR image. Now, what the exposure that you actually render at will most likely affect your tonemapped result, but it still does not necessitate more than one hdr image rendered from V-Ray.
Damien,
Yeah i thought that I’ve missed something from the ‘one vray image is enough’.
Well i tried to save the render as HDR (32b bit) file.
opened it in PS and converted it to three 8 bit images with different exposures.
Here’s what I’ve got: Img No’ 4
For comparison here’s the 3 times rendered image from before Img No’ 2
It turned pretty close (notice the wall detail/tones, and the noise at the sky)
Maybe it’s something that i did with the V-ray hdr raw file exposing/conversion.
We’re getting closer, but I still don’t think you’re getting it. You DO NOT have to split the image into 3 exposures AT ALL. You have you’re rendering as an HDR and that’s it. Saving out as 3 different exposures in PS does nothing except take an HDR, peel it apart, then put it back together again. All that matters for the tone mapping process is an HDR image…it doesn’t matter how you get it, just that its an HDR and has good dynamic range. If you’ve got the HDR from V-Ray with all the information that it will ever have, then there’s not need to do anything else to it other than let the TMO do its job.
Ok.
The testing continues:
Opened the V-Ray’s hdr image directly with photomatix, and did tone mapping.
The good news is that i used one v-ray image, and the less good is that it still doesn’t look good as Img No’ 2
As I understand it, Photomatix is first and foremost made for combining multiple exposure 8 bit images into 32 bit HDRIs. While the topic title sounds like your aim is to convert VRay output into an HDRI (either via actual 32 bit export or via multiple 8 bit exports converted to 32 bit in Photomatix), your ultimate goal seems to be to create an LDRI which looks like the “HDRI” photos you see on photographer’s websites and Flickr which of course are actually LDRIs created from HDRIs created from multiple LDR photographs.
My question is why bother with Photomatix at all? If you have a 32 bit HDRI from VRay, then just open it in Photoshop or Gimp and you can edit levels, gamma, offset, exposure or whatever until it looks the way you want, then convert it to 8 bit; it’ll give you way more control than Photomatix.
The first image below is straight 8 bit output from VRay, the second was exported as 32 bit and gamma, offset and exposure adjusted in PS before conversion to 8 bit. Of course (aside from both being extremely noisy due to very low render settings) the second is oversaturated, but the point is you have control over pretty much everything when you export straight out at 32 bit and no need for using extra programs other than PS.
Hey Jackson,
Well, actually photomatix exports the image to 8 bit image (or 16 bit), and it’s goal is to show (make it look like) high dynamics range in LDR image, like 8 bit.
I understand that Ps has HDR conversion, but i couldn’t get close to output like with photomatix.
As for the post’s title, you right, it’s not accurate. It should be more like: V-ray Raw Image to LDR Image (looking like HDRI) .
Witch HDR conversion method did you use?
Can do a test on the same scene as mine, so we’ll have a good comparison?
Here’s the link to Skp. file: Download
Sorry to flatly contradict you, but Photomatix is specifically designed to create 32 bit HDRIs from multiple differently-exposed 8 bit photographs - their homepage makes that pretty clear: http://www.hdrsoft.com/ It can export out to 8 bit, but it natively saves as 32 bit. PS can also do this, but personally I prefer the Photomatix workflow for combining LDRs into HDR format while using PS to do any HDR editing.
I just changed the image from 32 bit to 8 bit under Image>Mode in PS without altering Exposure or Gamma (I’d already adjusted these before).
Could you just post a link to your exported VRay HDRI render so I can have a look at it in PS?
I meant that that’s the goal of Photomatix for me (my use of it). Sorry for not being clear on it, I know what it’s capable of. Link to vray hdr output
ihave done also the same thing before using 5 podium render images and i share it at sketchucation. i have also used photomatix to create a hdri effect and truly so the effect is unusual. the thing even in using photograph i prefer photomatix tone mapping sliders..
on the other hand the effect is not as noticeable as using photograph and this perhaps the point of damien with rendered vray. nevertheless i found it as another way to explore post processing.